Member Avatar for rasizzle

Hey all,

Here is what I'm trying to do. I have a list whose output is:

[['4/18/94', '29.125', '442.46'], ['4/19/94', '29.336', '442.54'], ['1/20/04', '75.175', '1138.77'], ['1/21/04', '75.711', '1147.62'], ['1/22/04', '75.595', '1143.94']]

Item [1] in each list, the middle item, is the number I want to focus on. Currently, I have my sentiment range set between 30 and 75. These are sentiment numbers in item[1] of each list within the larger list.

I want to make it so that no 2 item[1]'s are below 30 in a row, and no 2 item[1]'s are above 75 in a row.

So looking only at the sentiment numbers, here is how I want my list ordered:

[['4/18/94', '29.125', '442.46'], ['1/20/04', '75.175', '1138.77']]

That's a simple version. But basically for every list within a list that has a sentiment number below 30, the next list within the list should have a sentiment number above 75.

for x in range(1,len(list_of_sentiments)):
    if float(list_of_sentiments[x][1]) < Low_Sentiment:
        if float(list_of_sentiments[x-1][1]) < Low_Sentiment:
            del list_of_sentiments[x]
print list_of_sentiments

I keep getting an index range error because of the x-1. I get the same error when i try to switch it around with an x+1. Is there any easier way to get rid of the lists with sentiment numbers in the same range as the item before?

  • 2 Contributors
  • 2 Replies
  • 118 Views
  • 6 Hours Discussion Span
  • Latest Post Latest Post by rasizzle

Dani AI

Generated

The IndexError you hit comes from mutating the list while iterating over an index range computed up front: removing items shortens the list and later indices go past the new end. Building a new list avoids that class of bug — which is why ’s append-into-result approach works reliably.

Two alternative, safe patterns:

  • In-place deletion but iterate backwards. Walking indices from the end to the start avoids shifting problems because removing data[i] does not affect indices you have yet to visit:

    for i in range(len(data)-1, 0, -1):
    curr = float(data[i][1])
    prev = float(data[i-1][1])
    if curr < LOW and prev < LOW:
        del data[i]
    elif curr > HIGH and prev > HIGH:
        del data[i]

  • Collapse consecutive runs with itertools.groupby and keep one representative per run (useful if you want to compress long runs of lows or highs to a single entry):

    from itertools import groupby

def category(item):
    n = float(item[1])
    return 'L' if n < LOW else 'H' if n > HIGH else 'M'

collapsed = []
for key, group in groupby(data, key=category):
    if key in ('L', 'H'):
        collapsed.append(next(group))   # keep first of each low/high run
    else:
        collapsed.extend(list(group))   # keep mids as-is (adjust as desired)

Small practical notes: decide how to treat boundary values (30 and 75) — use </<= consistently. Convert the middle-item strings to floats once up front to avoid repeated conversions and potential ValueError on malformed input. For very large streams prefer a generator-based approach so you don’t hold two full copies in memory. Finally, add a few unit tests that cover runs (LLL, HHH), alternating extremes (L,H,L), and mixed mids to confirm the behavior matches expectations.

Recommended Answers

All 2 Replies

Member Avatar for shadwickman

Got it :D

data = [['4/18/94', '29.125', '442.46'],
    ['4/19/94', '29.336', '442.54'],
    ['1/20/04', '75.175', '1138.77'],
    ['1/21/04', '75.711', '1147.62'],
    ['1/22/04', '75.595', '1143.94']]
    
result = []

for item in data:
    # current middle index as a float
    n = float(item[1])
    
    # if there are no items in the result list yet
    if not result:
        result.append(item)
        
    # otherwise, if there are,
    else:
        # previous number (last one in result list) as a float
        prevnum = float(result[-1][1])
        
        if (prevnum < 30 and n > 75) or (prevnum > 75 and n < 30):
            result.append(item)

"""
My result:
[['4/18/94', '29.125', '442.46'], ['1/20/04', '75.175', '1138.77']]
"""
commented: excellent advice +1
Member Avatar for rasizzle

Yes! That works perfectly. Thanks for coming through again. If you haven't guessed, I still haven't been able to fully use my Buy and Sell program. I am re-attempting to find a simpler way to do it and this looks great. Thanks again!

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.